By Maayan Hoffman
health & wellness 16
Artificial intelligence (AI) is on fire in the marketing, transportation and finance industries, but the healthcare arena has been slow to adopt this innovation. No more. Today, with the growth of clinical data captured effectively by digital health records, the medical world is increasingly taking advantage of AI tools. TAU scientists and their hospitalbased physician colleagues are leveraging new AI techniques such as machine learning and deep learning to accelerate, deepen and extend the impact of laboratory findings. The difference between regular data-crunching solutions and current AI systems is that the latter mimic human learning and teach themselves to make sense of vast amounts of information. As new patterns are “learned” and catalogued, the AI-driven software can ultimately assist and improve the performance of human care providers. Last year, TAU upped its game by inaugurating the Yandex Initiative in Machine Learning (ML), which supports mostly ML courses at the Blavatnik School of Computer Science, creating the next generation of leaders in industry. More nuanced diagnoses According to Talma Hendler, professor of psychiatry at TAU’s Sackler Faculty of Medicine and Sagol School of Neuroscience, AI software can help doctors gain insights into a patient’s clinical results in real or rapid time. With AI, a computation based on clinical data that might have taken a month in the past can now be computed in an hour. Increased AI will also allow for more personalized medical treatment and give doctors the ability to look at patients more holistically. If in the past, a patient would have been labeled as diabetic or having heart disease, “machine learning will cluster all of a patient’s ailments and perhaps help develop new, more nuanced correlations and labels.”
Hendler believes that, just as today every patient entering an emergency room is automatically given a blood test, in the future, some form of AI-related testing and consideration will become standard practice. In her own research, Hendler is working on a multidisciplinary platform to address mental health problems combining AI with advanced brain imaging and other neuroscience techniques. Specifically, she has been mapping objective brain markers and
Intelligent Enough to Make You Healthy? TAU biomedical researchers bring innovative AI techniques into the real-world clinical environment behavioral characteristics of individuals who are experiencing a mental disorder. The goal is that when a patient says he is depressed, she would be able to confirm this depression not only subjectively (by talking with the patient), but through objective factors (by examining his neural networks). “Being able to calculate a patient’s brain network functioning will help doctors formulate a personalized treatment plan,” Hendler said.
Genetic genius Hendler’s methodology is not unlike the work being done by the Sackler Faculty of Medicine’s Prof. Lina Basel Salmon, who heads the genetics institute at the TAU-affiliated Rabin Medical Center. She has been using deep learning and computer imaging techniques to train software to recognize genetic disorders among children based on their facial abnormalities.
“Nowadays there are so many sources of information,” she said, which can make it difficult to determine a diagnosis. Salmon helped spearhead a now Boston-based startup called FDNA and its Face2Gene app, which allows doctors to take a picture of a baby or toddler’s face and upload it to the app. Then, it provides the top 10 most likely genetic syndromes based on facial patterns. Doctors then help the technology narrow down potential diagnoses by uploading the patient’s symptoms to the app as well. Salmon said the software is about 91% accurate – much more accurate than even the best genetic specialists. She noted that the app is not meant to replace the physician, but rather to provide support. “This will help the physician extract the best and most precise information and suggest to him or her ideas for treatment and management,” she said.
